Wiki Knowledge Ingestor
Automates the processing of raw notes into a structured, interconnected Obsidian-based knowledge base.
The ingest_raw skill transforms Claude into an intelligent wiki maintainer by systematically organizing raw information into a structured digital garden. It scans unprocessed files, prioritizes them based on context, and synthesizes content into dedicated entity, concept, and source pages. Beyond simple ingestion, it manages complex bidirectional linking, updates frontmatter metadata for relationship tracking, and maintains a detailed activity log, ensuring your personal knowledge base remains consistent, cited, and easy to navigate.
主な機能
01Review Workflow: Marks ingested files as pending_review with detailed processing timestamps for manual verification.
02112 GitHub stars
03Metadata Management: Updates YAML frontmatter for relationships like 'supports', 'contradicts', and 'related_sources'.
04Conflict Resolution: Explicitly flags contradictions between new sources and existing wiki content instead of silent overwriting.
05Automated Raw Data Processing: Prioritizes and ingest files from unprocessed directories into the wiki structure.
06Smart Linking & Synthesis: Automatically creates bidirectional links and evidence-based nodes for concepts and entities.
ユースケース
01Building a local LLM-assisted knowledge base that tracks source evidence for every claim.
02Transforming a pile of raw research notes into a structured, cited Obsidian wiki.
03Maintaining an automated digital garden where new information is cross-referenced with existing concepts.